Absolute counts versus relative abundances in A. thaliana leaf bacterial networks

Plant microbial communities, called microbiomes, affect host productivity, defense responses, and fitness. As a consequence, researchers aim to tease apart interactions among hosts, microbes and the abiotic environment to develop means of using microbiomes to improve crop productivity in the field. For these experiments, we need effective methods of characterizing the microbiome within plant leaves, as well as effects on host health. Canonical methods for characterizing the microbiome use relative abundances of sequenced marker genes to generate community profiles, but we can increase our statistical power by using absolute counts of microbes. Deriving absolute counts requires extensive sequencing because there is currently no effective microbial DNA enrichment step, and therefore host DNA fragments inundate the reads. Here we use deep sequencing to derive absolute microbial counts from the leaves of A. thaliana grown in the presence of microbes in a field site in Sweden. We also test the efficacy of two potential methods of microbial enrichment for future research and examine the introduced biases in the characterization of the microbial community. We use these results to generate networks of inferred microbe-microbe interactions and compare these networks to previous networks derived from relative abundance sequencing. Finally, we briefly describe testing the importance of key members of the community in A. thaliana greenhouse experiments